Technology DevelopmentTechnology Development

Under Prof. Chen's conceptual guidance and new web interface design and development, Global Memory Net(GMNet)'s in-house LINUX/MySQL/PHP-based Integrated Multimedia Content Retrieval System (i-M-C-S) has been developed with substantial contributions from our international visiting researchers. Specifically Shengqiang Zhang, University of Sichuan, China developed the early Version 2.0 during his 5-month visit at Simmons in the second half of 2005. His work was further developed and greatly refined by Boris Badurina, University J. J. Strossmayer, Osijek, Croatia and Xiaobo Xiao, Xian Jiaotong University, China during their 3-month visit at Simmons in the second half of 2005 (Versions 3.0 and 4.0).

The system for GMNet has been developed at Simmons College under the sponsorship of the US National Science Foundation between 2001 and August 2010. Since September 2010 until the completion of this project, the system development and modification has been supported by Global Connection and Collaboration.

After Boris Badurina's return to Croatia in 2005, he has continued to be closely involved with the system development of GMNet. We are currently at Version 5.0, and are developing several generic versions for adaptation and use by other interested organization.

The conceptual challenges for developing our user-friendly world digital image library are many. The multi-purpose, multi-collection and multilingual search capabilities of our i-M-C-S system include the basic traditional textual retrieval by metadata fields as well as the random exploration of a chosen image collection when its content is not familiar to the user. When a desired image is located, users can request all images of similar color and shape by selecting "Similar." This feature uses the cutting-edge content-based image retrieval (CBIR), SIMPLIcity, developed by Profs. James Z. Wang and Li Jia at Stanford (1998-2000) and since then at the Penn State University (PSU). The user can also enlarge and browse portions of the image with dynamically generated digital water mark for copyright and IP protection, as well as find related metadata information. Links to relevant web sites are provided when available, and further bibliographical and multimedia searches can also be instantly provided when available. Geographical access to image collections is also provided. More descriptions of the system features can be found in the graphics of the project brochure. In addition, content management tools as well as user registration and upload functions are also developed.

Technical Development at PSU - Prior to April 2005, GMNet Version 1.0 was developed for internal R&D use. Its limited information retrieval capabilities were developed by Prof. James Z. Wang of PSU. It linked the search display to the HTML-based web page developed by the Simmons Team. The text retrieval engine of PSU searches the metadata fields of the databases initially created at Simmons using MS Access. The visual similarity search system uses the SIMPLIcity image retrieval engine. We are grateful to Prof. Wang for his effort and collaboration.

Latest R&D Project - International Collaboration to Advance User-oriented Technologies for Managing and distributing Images in Digital Libraries," NSF/IIS-0333036 (PI: Ching-chih Chen, Co-PIs: James Z. Wang and Jianbo Shi, September 2004-August 2010). The project is jointly conducted by Simmons College, The Pennsylvania State University, and University of Pennsylvania. Results of this project have been incorporated as part of GMNet.